Fault diagnosis for diesel engine based on immune wavelet neural network

This paper proposes the wavelet neural network (WNN) based on clonal selection algorithm (CLONALG) for using in fault diagnosis of marine diesel engine. CLONALG initializes the WNN´s weights and biases, the ergodic weights and biases are used for further net-training. The fault diagnosis for marine diesel engine is conducted by using the well-trained wavelet network, in order to illustrate the performance of this model. The results obtained indicate that the WNN based on CLONALG can avoid the local extremum, and the convergence, generalization and the capability of fault diagnosis are all improved.